optimus-ep 0.6.5 → 0.6.9

data/History.txt

@@ -1,3 +1,16 @@
+== 0.6.9 2/25/2008
+* New features:
+  * Supports parsing raw tab-delmimted files
+  * Allows Procs as starting values for ComputedColumns
+* Bug fixes:
+  * No longer jumbles the order of Eprime log files
+
+== 0.6.5 9/2/2008
+* New features:
+  * Speed! Speed! Speed! Should be something like 10x faster.
+  * Procs are now allowed in ComputedColumns -- so if you want to do really
+    arbitrary things in your column computation, have at it.
+
 == 0.6.0 7/14/2008
 * New features:
   * Added extract_timings, a script to pull stimulus timing data from eprime

data/Rakefile

@@ -1,6 +1,7 @@
 require 'rake'
 require 'spec/rake/spectask'
-require 'config/requirements'
+#require 'config/requirements'
+require 'hoe'
 require 'config/hoe' # setup Hoe + all gem configuration
 
 Dir['tasks/**/*.rake'].each { |rake| load rake }

data/lib/eprime_data.rb

@@ -115,10 +115,6 @@ module Eprime
       return row
     end
     
-    #def add_row_values(values, sort_value = 1)
-    #  r = Row.new(self, values, sort_value)
-    #end
-    
     def find_column_index(col_id)
       @column_hash[col_id]
     end
@@ -126,8 +122,8 @@ module Eprime
     def find_or_add_column_index(col_id)
       index_id = find_column_index(col_id)
       # If index_id was a string, nil means we may want to add it. If it's a
-      # numeric index, we want to return nil from here -- we're not gonna add unnamed
-      # indexes.
+      # numeric index, we want to return nil from here -- we're not gonna add 
+      # unnamed indexes.
       return index_id if index_id or col_id.is_a?(Fixnum)
       # In this case, we're adding a column...
       index = @columns.size
@@ -135,7 +131,10 @@ module Eprime
       @column_hash[col_id] = index
       @column_hash[index] = index
       if @columns_set_in_initialize and not @options[:ignore_warnings]
-        raise ColumnAddedWarning.new("Error: Added column #{col_id} after specifying columns at init", index)
+        raise ColumnAddedWarning.new(
+          "Error: Added column #{col_id} after specifying columns at init", 
+          index
+        )
       end
       return index
     end
@@ -159,7 +158,7 @@ module Eprime
       
       def [](index)
         num_index = @parent.find_column_index(index)
-        unless (num_index.is_a?(Fixnum) and @parent.columns.length > num_index)
+        unless (num_index.is_a?(Fixnum) and @parent.columns.length>num_index)
           raise IndexError.new("Column #{num_index} does not exist")
         end
         return @data[num_index]

data/lib/log_file_parser.rb

@@ -26,8 +26,6 @@ module Eprime
       attr_reader :frames
       attr_reader :levels
       attr_reader :top_level
-      attr_reader :skip_columns
-      
       # Valid things for the options hash:
       #   :columns => an array of strings, predefining the expected columns 
       #               (and their order)
@@ -39,8 +37,9 @@ module Eprime
         @force = options[:force]
         @file = file
         @levels = [''] # The 0 index should be blank.
-        @top_level = 0 # This is the level of the frame that'll generate output rows
-        @skip_columns = {} # A hash of columns we *don't* want to add -- just define the strings
+        @top_level = 0 # This is the level of the frame that'll 
+                       # generate output rows
+        @found_cols = ColumnList.new()
       end
       
       def make_frames!
@@ -56,30 +55,17 @@ module Eprime
             make_frames!
           end
         rescue Exception => e
-          unless @force
-            raise e
-          end
-        end
-        if @columns
-          data = Eprime::Data.new(@columns)
-        else
-          data = Eprime::Data.new
+          raise e unless @force
         end
+    
+        @columns ||= @found_cols.names
+        data = Eprime::Data.new(@columns)
         self.top_frames.each do |frame|
           row = data.add_row
-          frame.columns.each do |column, value|
-            begin
-              # Do a check for columns to skip -- this will happen in the case
-              # where you have Procedure[Session] and Procedure[Task] -- we
-              # shouldn't have Procedure, in that case.
-              unless @skip_columns[column]
-                row[column] = value
-              end
-            rescue Exception => e
-              unless @force
-                raise e
-              end
-            end
+          @found_cols.names_with_cols.each do |pair|
+            name, col = *pair
+            val = frame.get(col)
+            row[name] = val
           end
         end
         return data
@@ -105,8 +91,9 @@ module Eprime
       def frameify(file)
         in_frame = false
         frames = []
+        # This 
         frame = Frame.new(self)
-        level = 0
+        frame.level = 0
         file.each_line do |line|
           # TODO? Refactor this out into its own private function
           l_s = line.strip
@@ -129,10 +116,15 @@ module Eprime
               # One more special thing: Experiment gets renamed ExperimentName. WTF?
               key = "ExperimentName" if key == "Experiment"
               frame[key] = val
+              @found_cols.store(Column.new(key, frame.level))
             end
           end
         end
-        raise DamagedFileError.new("Last frame never closed in #{file.path}") if in_frame
+        if in_frame
+          raise DamagedFileError.new(
+            "Last frame never closed in #{file.path}"
+          ) 
+          end
         return frames
       end
       
@@ -148,6 +140,7 @@ module Eprime
             end
           else
             if key == HEADER_END
+              @found_cols.levels = @levels
               file.rewind
               return # Get out of this function!
             else
@@ -164,6 +157,7 @@ module Eprime
         @frames.each do |frame|
           counts[frame.level] += 1
           key = @levels[frame.level]
+          @found_cols.store(Column.new(key, frame.level))
           frame[key] = counts[frame.level]
           counts.fill(0, (frame.level+1)..@levels.length)
         end
@@ -178,6 +172,8 @@ module Eprime
       end
       
       class Frame
+        include Enumerable
+        
         attr_accessor :level
         attr_accessor :parent
         def initialize(parser)
@@ -187,25 +183,99 @@ module Eprime
           @parser = parser
         end
         
-        def columns
-          my_data = @data.dup
-          return my_data if @parent.nil?
-          parent_data = @parent.columns
-          parent_data.each do |key, val|
-            if my_data.has_key?(key)
-              @parser.skip_column(key)
-              # Append a string like "[Session]" or "[Block]" to the key name
-              my_data["#{key}[#{@parser.levels[@level]}]"] = my_data[key]
-              my_data["#{key}[#{@parser.levels[@parent.level]}]"] = val
-            else
-              my_data[key] = parent_data[key]
-            end
+        # Methods to make this behave hashlike. Don't just delegate to 
+        # the @data hash; that's less clear.
+        def [](key)
+          return @data[Column.new(key, @level).to_s]
+        end
+        
+        def []=(key, val)
+          @data[Column.new(key, @level).to_s] = val
+        end
+        
+        def keys
+          @data.keys
+        end
+        
+        def each
+          @data.each do |k, v|
+            yield k, v
+          end
+        end
+        
+        def get(col)
+          # If the value is supposed to be at our level, return it (nil is OK)
+          return @data[col.to_s] if col.level == @level
+          # If it could be in our parent, return that.
+          return @parent.get(col) if (@parent && col.level < @level)
+          # If that's not an option,
+          return nil
+        end
+      end
+      
+      class ColumnList
+        attr_accessor :levels
+        def initialize(levels = [], cols = [])
+          @levels = levels
+          @cols = cols
+          @name_uses = Hash.new(0)
+        end
+        
+        def store(col)
+          if (col.level >= @levels.length  or col.level < 1)
+            raise IndexError.new(
+              "Level #{col.level} must be between 1 and #{@levels.length-1}")
+          end
+          if not @cols.include?(col)
+            @cols << col
+            @name_uses[col.name] += 1
           end
-          return my_data
         end
         
-        def method_missing(meth, *args)
-          @data.send meth, *args
+        def names
+          return self.names_with_cols.map { |c| 
+            c[0]
+          }
+        end
+        
+        def names_with_cols
+          ncm = sorted_cols.map {|c| [
+              (@name_uses[c.name]==1) ? 
+                c.name : "#{c.name}[#{@levels[c.level]}]",
+              c ]}
+          return ncm
+        end
+        
+        def sorted_cols
+          cwi = []
+          @cols.each_with_index do |col, i|
+            cwi << [col, i]
+          end
+          return cwi.sort_by {|elem| [elem[0].level, elem[1]]}.map {|elem| 
+            elem[0]
+          }
+        end
+        
+
+      end
+
+      class Column
+        attr_accessor :name, :level
+        def initialize(name, level)
+          @name = name
+          @level = level
+        end
+        
+        def ==(c)
+          @name == c.name and @level == c.level
+        end
+        
+        def hash
+          return self.to_s.hash
+        end
+        
+        def to_s
+          "#{@name}[#{@level}]"
         end
       end
     end # Class LogfileParser

data/lib/transformers/column_calculator.rb

@@ -240,7 +240,7 @@ module Eprime
           end
           
           # Allow defining the column computation as a lambda
-          return @expression.expr.call(row) if @expression.expr.is_a? Proc
+          return @expression.expr.call(row) if @expression.expr.respond_to? :call
             
 
           column_names = @expression.columns
@@ -278,9 +278,12 @@ module Eprime
           if @reset_when.call(row)
             @current_value = @start_value
           end
+          if @current_value.respond_to? :call
+            @current_value = @current_value.call(row)
+          end
         
           if @count_when.call(row)
-            if @count_by.is_a? Proc
+            if @count_by.respond_to? :call
               @current_value = @count_by.call(@current_value)
             elsif @count_by.is_a?(Symbol) || @count_by.is_a?(String)
               @current_value = @current_value.send(@count_by)

data/lib/version.rb

@@ -2,7 +2,7 @@ module Eprime
   module VERSION
     MAJOR = 0
     MINOR = 6
-    TINY  = 5
+    TINY  = 9
     
     STRING = [MAJOR, MINOR, TINY].join('.')
   end

data/spec/log_file_parser_spec.rb

@@ -66,7 +66,7 @@ describe Eprime::Reader::LogfileParser do
     it "should have a parent in the first frame" do
       @reader.frames.first.parent.should_not be_nil
     end
-  
+    
     describe "making eprime data" do
       before :each do
         @eprime = @reader.to_eprime
@@ -76,6 +76,11 @@ describe Eprime::Reader::LogfileParser do
         @eprime.length.should == 3
       end
       
+      it "should follow the column order in the example file" do
+        @eprime.columns[0].should == "ExperimentName"
+        @eprime.columns[1].should == "SessionDate"
+      end
+      
       it "should ignore extra colons in input data" do
         @eprime.first['SessionTime'].should == '11:11:11'
       end
@@ -88,10 +93,6 @@ describe Eprime::Reader::LogfileParser do
         @eprime.columns.should_not include("CarriedVal")
       end
       
-      it "should mark ambiguous columns for skip" do
-        @reader.skip_columns.should include("CarriedVal")
-      end
-    
       it "should include columns from level 2 and level 1 frames" do
         @eprime.columns.should include("RandomSeed")
         @eprime.columns.should include("BlockTitle")
@@ -153,4 +154,57 @@ describe Eprime::Reader::LogfileParser do
     
   end
   
+end
+
+describe Eprime::Reader::LogfileParser::ColumnList do
+  before :each do
+    @cklass = Eprime::Reader::LogfileParser::Column
+    @levels = ['', 'Foo', 'Bar']
+    @list = Eprime::Reader::LogfileParser::ColumnList.new(@levels)
+  end
+  
+  it "should raise error when storing index 0" do
+    lambda {
+      @list.store(@cklass.new('test', 0))
+    }.should raise_error(IndexError)
+  end
+  
+  it "should raise error when storing out of bounds" do
+    lambda {
+      @list.store(@cklass.new('test', @levels.length))
+    }.should raise_error(IndexError)
+  end
+    
+  it "should record and return names" do
+    @list.store(@cklass.new('test', 1))
+    @list.names.should == ['test']
+  end
+  
+  it "should not re-add repeated name at same level" do
+    @list.store(@cklass.new('test', 1))
+    @list.store(@cklass.new('test', 1))
+    @list.names.should == ['test']
+  end
+
+  it "should record unique column names at different levels" do
+    @list.store(@cklass.new('test', 1))
+    @list.store(@cklass.new('another_test', 2))
+    @list.names.should == ['test', 'another_test']
+  end
+  
+  it "should add level names to repeated column names at different levels" do
+    @list.store(@cklass.new('test', 1))
+    @list.store(@cklass.new('test', 2))
+    @list.names.should == ['test[Foo]', 'test[Bar]']
+  end
+  
+  it "should return paired columns with names" do
+    col = @cklass.new('test', 1)
+    @list.store(col)
+    
+    @list.names_with_cols.should == [
+      ['test', col]
+    ]
+  end
+  
 end

data/spec/samples/optimus_log.txt

@@ -94,7 +94,6 @@ Display.RefreshRate: 60.000
 	*** LogFrame End ***
 Level: 1
 *** LogFrame Start ***
-CarriedVal: SessionLevel
 Experiment: optimus_test
 SessionDate: 03-15-2008
 SessionTime: 11:11:11
@@ -105,6 +104,7 @@ Session: 1
 Display.RefreshRate: 60.000
 Clock.Scale: 1
 NumPeriods: 3 
+CarriedVal: SessionLevel
 PeriodA: 10000 
 PeriodB: 10000 
 *** LogFrame End ***

data/spec/transformers/column_calculator_spec.rb

@@ -395,6 +395,18 @@ describe Eprime::Transformers::ColumnCalculator do
         row['reset_on_sparse'].should == i
       end
     end
+    
+    it "should accept a Proc as a :start_value" do
+      @calc.counter_column "scan_delay", :start_value => lambda {|row|
+        row['stim_time'].to_i - row['run_start'].to_i
+      }
+      sd = @calc[0]['scan_delay']
+      i = 0
+      @calc.each do |row|
+        row['scan_delay'].should == sd+i
+        i += 1
+      end
+    end
   end
 end
 

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification 
 name: optimus-ep
 version: !ruby/object:Gem::Version 
-  version: 0.6.5
+  version: 0.6.9
 platform: ruby
 authors: 
 - Nathan Vack
@@ -9,7 +9,7 @@ autorequire:
 bindir: bin
 cert_chain: []
 
-date: 2008-09-02 00:00:00 -05:00
+date: 2009-02-25 00:00:00 -06:00
 default_executable: 
 dependencies: 
 - !ruby/object:Gem::Dependency 
@@ -30,7 +30,7 @@ dependencies:
     requirements: 
     - - ">="
       - !ruby/object:Gem::Version 
-        version: 1.7.0
+        version: 1.8.2
     version: 
 description: A collection of utilities for working with Eprime data files
 email: 
@@ -131,7 +131,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
 requirements: []
 
 rubyforge_project: optimus-ep
-rubygems_version: 1.2.0
+rubygems_version: 1.3.1
 signing_key: 
 specification_version: 2
 summary: A collection of utilities for working with Eprime data files